Person-Level Viewership Projection via Demographic Attribution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for estimating television viewership struggle with accurately projecting person-level viewership from household-level data, as household-level data lacks information on individual viewers, leading to inaccuracies in unique viewer numbers and viewing minutes, and person-level data has a small sample size with high margins of error.
Innovation Solution
The approach involves accessing tuning data and panelist viewing data to determine fractional viewership values and person-level minutes by matching demographic profiles, allowing for the projection of person-level viewership from household-level data through demographic attribution, which integrates both data sets to improve accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If household-level tuning data is used for viewership estimation, then the sample size is large, but the data lacks information on individual viewers leading to inaccurate unique viewer numbers and viewing minutes
Solution Approach 1:
The patent uses demographic profiles as an intermediary to bridge household-level tuning data and person-level viewership estimation. By matching panelist demographic profiles with household member demographics, the system can attribute viewing behavior to individual demographic groups without directly observing individual panelists in every household, thus recovering individual viewer information from aggregate household data.
Solution Approach 2:
The patent creates a demographic copy of the household composition using panelist data. By constructing demographic profiles that mirror the household's member demographics and applying panelist viewing patterns to these profiles, the system generates person-level viewership estimates that replicate what would be observed if actual individual viewing data were available.
2Loss of information
If person-level panelist viewing data is used for viewership estimation, then individual viewer information is available, but the sample size is small with high margins of error
Solution Approach 1:
The patent merges household-level tuning data with panelist viewing data through demographic attribution. By combining the large sample size of household tuning events with the detailed individual viewer information from panelists (matched through demographic profiles), the system achieves both adequate sample size and precise individual viewer measurement, reducing margins of error.
3Measurement precision
If demographic attribution is applied to project person-level viewership, then accuracy of audience measurement is improved, but the complexity of data processing increases
Solution Approach 1:
The patent performs preliminary demographic profiling and matching before the actual viewership attribution process. By pre-establishing demographic profiles for households and pre-matching panelists to household members based on demographic criteria, the system simplifies the subsequent viewership projection process, making the complex demographic attribution more manageable and efficient.
Data Source
AI summary
The present disclosure describes techniques for projecting person-level viewership from household-level tuning events. One example method includes accessing tuning data representing television tuning events associated with particular households; accessing panelist viewing data representing television viewing events associated with panelists; and for at least one tuning event represented by the tuning data: accessing household member data associated with the tuning event; determining a member minutes value associated with each individual member of the particular household for the tuning event; and determining a fractional viewership value for each individual member of the particular household associated with the tuning event.